Domaines
Statistical physics
Physics of living systems
Type of internship
Théorique, numérique Description
Why do some images feel visually compelling while others appear dull, chaotic or artificial? This internship will investigate whether aesthetic preference can be related to measurable statistical properties of images. Building on our previous work on structural complexity, multiscale relevance, and quantitative color harmony, we will combine large human-preference datasets with interpretable descriptors of image organisation to identify which visual structures are consistently associated with appeal.
Contact
Michael Benzaquen